Asymmetric Single Graft Block Copolymers: Effect of Molecular

Macromolecules 0 (proofing), .... Polymer Chemistry 2011 2, 69 ... Science Part A: Polymer Chemistry 2001 39 (10.1002/1099-0518(20010101)39:11.0...
2 downloads 0 Views 264KB Size
3732

Macromolecules 1997, 30, 3732-3738

Asymmetric Single Graft Block Copolymers: Effect of Molecular Architecture on Morphology Chin Lee and Samuel P. Gido* Polymer Science and Engineering Department, W. M. Keck Electron Microscopy Laboratory, University of Massachusetts, Amherst, Massachusetts 01003

Marinos Pitsikalis and Jimmy W. Mays Department of Chemistry, University of Alabama at Birmingham, Birmingham, Alabama 35294

Nora Beck Tan,† Samuel F. Trevino,† and Nikos Hadjichristidis‡ Polymers Research Branch, U.S. Army Research Laboratory, APG, Maryland 211005-5069, Reactor Radiation Division, NIST, Gaithersburg, Maryland 20899, Department of Chemistry, University of Athens, Panepistimiopolis Zografou 15771, Athens, Greece, and Institute of Electronic Structure and Laser, 711 10 Heraklion, Crete, Greece Received February 5, 1997; Revised Manuscript Received April 15, 1997X

ABSTRACT: This paper reports on the synthesis and morphological characterization of graft block copolymers in which a single polystyrene (PS) arm was grafted at an asymmetric position along a polyisoprene (PI) backbone. These materials represent a model series of asymmetric simple graft (ASG) block copolymer structures. The synthesis of these materials was carried out with methods developed for three-arm “miktoarm” star copolymers using anionic polymerization high-vacuum techniques with cholorosilane linking agents. The three arms were two polyisoprene blocks with different degrees of polymerization and one deuterated polystyrene block, which formed an asymmetric simple graft structure (ASG). Molecular characterization was performed using size exclusion chromatography (SEC) with refractive index and UV detection, membrane osmometry, and low-angle laser light scattering. These techniques confirmed that the materials exhibited narrow molecular weight distributions and low compositional heterogeneity. The morphologies formed by these samples were characterized using transmission electron microscopy (TEM) and small-angle neutron scattering (SANS). The ASG structures were compared with structures formed by linear diblock copolymers and other samples having miktoarm structures, like I2S (symmetric simple graft with PI backbone and one PS branch grafted from the middle) and I3S (three equal PI arms and one PS arm). Comparisons of the morphologies formed and their dimensions indicated that the chain stretching and lateral crowding due to the miktoarm architecture was partially alleviated by the different lengths of PI blocks in ASG.

Introduction Diblock copolymer morphological behavior has been mapped by experiments and predicted by theory based on a two-parameter model utilizing volume fraction (φ), and χN, the product of the Flory χ parameter and degree of polymerization.1-6 In the strong segregation limit (large χN), the morphology formed is solely dictated by the relative volume fractions of the respective blocks. Recently, because of a successful strategy to synthesize well-defined miktoarm star block copolymers of the type AmBn developed by the group of Hadjichristidis, the molecular architecture was demonstrated experimentally as an additional variable for control of morphology.7-11 Connecting different numbers of A and B arms at one junction point (i.e. m not equal to n) creates an asymmetry, due to molecular architecture, in the way that the opposing A and B block materials fill space. Compared to a linear diblock, block materials with two or more arms experience more stretching and more lateral crowding. Additionally, if m and n are not equal, this extra chain stretching will be greater on the side of the interface with a larger number of arms per molecule. * To whom correspondence should be addressed. † U.S. Army Research Laboratory and NIST. ‡ University of Athens and Institute of Electronic Structure and Laser. X Abstract published in Advance ACS Abstracts, June 1, 1997.

S0024-9297(97)00172-1 CCC: $14.00

To alleviate this asymmetry in chain stretching, there is a shift in interfacial curvature preference: The interface tends to decrease in curvature when viewed from the side of the block material with higher arm number. If the higher arm number occurs on the concave side of the interface, the interface will tend to be less concave than it would be in a linear diblock of the same relative volume fraction. If the higher arm number occurs on the convex side of the interface, then the interfacial curvature will tend to be more convex than it would be in a linear diblock. These curvature preferences, as illustrated in a mean-field morphology diagram calculated by Milner (Figure 1), result in a shift of the morphological transition lines toward higher volume fractions of the lower arm number component. In these calculations, the asymmetry caused by the difference in A and B arm numbers was lumped with the conformational asymmetry inherent in the A and B polymer chains into one asymmetry factor, . Our previous study of a series of I2S materials (two polyisoprene arms with equal molecular weight and one polystyene arm) found good agreement with the calculated morphology diagram, except for one sample.10 The sample at 0.81 PS volume fraction formed a new randomly ordered, worm-like micelle structure, which was demonstrated to be an equilibrium structure by careful annealing experiments.9 Another series of I3S materials together with several S2I and I2S samples were studied by Tselikas et al.11 These materials © 1997 American Chemical Society

Macromolecules, Vol. 30, No. 13, 1997

Asymmetric Single Graft Block Copolymers 3733 (Aldrich) and lithium dispersion (Aldrich) under vacuum and was subsequently diluted with benzene to the concentrations needed for polymer synthesis. The procedure used for the synthesis of simple grafts, outlined below, is similar to the method adopted by Iatrou and Hadjichristidis12 for the synthesis of ABC miktoarm star terpolymers.

s-BuLi + Is f PIR-Li+ PIR-Li+ + CH3SiCl3 f CH3Si(PI)RCl2 + CH3SiCl3v + LiCl (excess) I dS + s-BuLi f PS-Li+ Figure 1. Theoretical phase diagram calculated by Milner. Each triangle represents a specific ASG sample corresponding to the number within the triangle. The oval represents an I2S sample from ref 11.

I + dPS-Li+ titration CH3Si(PI)R(dPS)Cl f II

showed that Milner’s calculation slightly overestimated the degree of shift in morphology boundaries. However, both experiment and theory have demonstrated that miktoarm architecture can have a profound effect on morphology. In the present paper, a series of three-arm “miktoarm” I2S block copolymers were synthesized utilizing chlorosilane coupling chemistry.12 One PI arm is about 3 times longer than the other PI arm. This corresponds to a simple graft structure in which τ ≈ 0.25, where τ is the fraction distance along the PI backbone at which the PS graft occurs. The morphology behavior of this asymmetric simple graft structure cannot be predicted directly by Milner’s morphology diagram, because the calculation of the  parameter assumes that all blocks of the same type are of the same molecular weight (i.e., τ ) 0.5 for the I2S materials studied previously). In order to generalize the morphology diagram to the asymmetric simple graft structure (ASG), we postulated the following type of relationship in our previous paper: 13  ) f(τ)(l /l )1/2. Here, f(τ) represents the asymmetry A B due to the architecture, and (lA/lB)1/2 expresses the conformational asymmetry between the two block materials. li (i ) A or B) is the ratio of segment volume to the square of statistical segment length for the i block. Since the lateral crowding and additional chain stretching associated with the two chains on one side of the interface in a symmetrically grafted structure (τ ) 0.5) should be partially alleviated in the asymmetric simple graft structure, f(τ) should be between 1 (linear diblock) and 2 (symmetric simple graft architecture, I2S). The morphology behavior of the five samples presented in this paper basically agreed with this assumption and allows us to place some basic constraints on the form of the f(τ) relationship.

II + PIβ-Li+ f CH3Si(PI)R(dPS)(PI)β (excess)

Experimental Section Synthesis of Asymmetric Simple Grafts. The asymmetric simple grafts were synthesized by high-vacuum, anionic polymerization in all-glass reactors, provided with breakseals and constrictions for addition of reagents and removal of samples, respectively.14 All reactors were purged with nbutyllithium and rinsed with benzene, which was the solvent for the reactions, prior to introduction of reactants. The purification of benzene (solvent), styrene-d8 and isoprene (monomers obtained from Cambridge Isotopes and Aldrich), trichloromethylsilane (linking agent, Aldrich), and methanol (terminating agent) to the standards required for anionic polymerization was performed using reported procedures.14 The initiator used for all polymerizations, secbutyllithium, was prepared by the reaction of sec-butyl chloride

s-BuLi + Is f PIβ-Li+

A dilute solution (ca. 2.5% w/v) of the PIR-Li+ solution was reacted with a very large excess of CH3SiCl3 (Cl/Li = 100) to avoid coupling. The excess chlorosilane and benzene were removed on the vacuum line. The end-functionalized polymer was then redissolved in pure benzene (distilled through the vacuum line), followed by removal of the solvent under vacuum conditions. After pumping for 1 day, the above procedure was repeated once more, and finally, the product was pumped down for several days at 4 °C. The duration was dependent upon the molecular weight of the sample and the quantity of the linking agent used. Size exclusion chromatography (SEC) analysis of the end-capped specimen (I) showed less than 2% coupling in all cases, usually less than 1%. In the second step, the dilute (ca. 2.5% w/v) living dPS-Li+ solution was added dropwise to a solution of the difunctional linking agent (I). This procedure, called “titration”,12 was monitored visually and via SEC analysis of samples removed from the reaction mixture. This methodology led to the formation of the diblock product with one remaining Si-Cl bond at the junction point (II). Finally, a small excess of living PIβ-Li+ (ca. 10% w/v), having a molecular weight about one-third that of PIR chains, was added to the reactor for the preparation of the final product. The excess of PIβ-Li+ was then deactivated with methanol. The product was purified by fractionation involving addition of methanol to dilute solutions (